Written by Zoe Berthold

Artificial intelligence (AI) is the manipulation of data to get information, particularly from large data sets. At the moment, AI is used for ‘narrow’ tasks; while it can perform with greater efficiency, an AI cannot mimic the complex processes by which the brain
functions. Radiology is the study and treatment of diseases using medical
imaging, using tools such as x-rays.

Artificial intelligence can be used to much better analyze images of different types of nodules, lesions, polyps, tumors, and more in radiological study. It is estimated that a radiologist needs to analyze approximately 3-4 images per second in an eight hour work week; the use of artificial intelligence could provide drastically needed improvements regarding efficiency. Within oncology imaging, AI can be used to detect lesions, characterize qualities like size and internal texture, and monitor treatment such as radiation to determine correct dosage.

Under this supposedly marvelously effective system, artificial intelligence would be trusted with vast amounts of data pertaining to persons medical health. Such an AI would need to be meticulously constructed and reviewed to verify their accuracy and comprehension. Ethical dilemmas also arise; with the technological handling of medical information, restricted databases could more easily be breached and the question or whether or not an artificial intelligence should be trusted with a person’s life becomes relevant. However, as the amount of data to evaluate increases and performance benchmarks are raised, artificial intelligence will undoubtedly become necessary in radiological study and surpass human capacity in its ability to process and understand data.

Citation:
Hosny, A., Parmar, C., Quackenbush, J., Schwartz, L. H., & Aerts, H. J. (2018). Artificial intelligence in radiology. Nature Reviews Cancer, 18(8), 500-510. doi:10.1038/s41568-018-0016-5